Exactly why many believe that one app “iNat Next” cannot please everyone, and why Seek should not be abandoned in the way many feel it has been.
Sorry to make you feel strange! I happen to live close enough to where iNat is based, and to be a gregarious enough person, to have met several of the people who work on it, often at a bioblizt or similar. I do not feel that having met these very personable people gives me much insight into the organization, or changes my overall experience of iNat.
My impression is that like most newish organizations, iNat is going through substantial growing pains. Usually these are hidden from the public, but @kueda has, for better or worse, told us some of what he’s seen behind it.
Hopeful for the elaboration of this from the December update:
The CV system had to be trained and it’s only trained by the community based identifications on observations that reach research grade.
This is not correct and it’s more complicated. Even casual observations can be used as training data depending on circumstance and if they meet the eligibility requirements.
I was happy with the survey. I had my pet peeve, too specific suggestion, listed.
No.
Anything that is streamed can be scrubbed and I am sure AI other than the CV are using them.
Can you explain what your concerns are?
For my part, I DO want my photos to inform the automated ID suggestions. I try to take photos that will show important aspects of mistletoes, both so that other iNat users can see what is distinctive about a species and so the Computer Vision will include those aspects in its recognition.
I am rather dismayed that several of the recent replies on this thread seem to be digressing onto various sideline topics. To bring this back. Why does it require another forum user to dig some text out from a different forum post (see above “updates for December”) to get indirect viewpoint that someone at the staff end might have been reading any of the above? This whole thread started at the end of a week, ran over the weekend, but now has continued over several working week days. Now we have a long thread (in part) about many forum users being concerned by a lack of response at staff end?
Wait, perhaps this:
“Hey everyone, thank you all for many interesting viewpoints, great discussion! Together you’ve all contributed so much over the years, we really value you! Without you all we would not have such a great community, nor be able to provide new users with the great value from our combined efforts and experience. We just put some context into our December update where we tell you about some recent developments, check them out and we will try to continue with monthly announcements to keep you all informed. In the new year we look forward to bringing in more features and updates to grow together stronger, but please know that our community matters to us!”
There you go iNat, I’ve just done a hypothetical general goodwill comment in-line with what I was really hoping to have seen by now.
Thanks for this, you’re right that it would have been helpful for us to update you all here as well. We have been reading through this (and other) forum threads and really do appreciate everyone’s feedback and contributions to iNaturalist. Thank you for being patient with us while we work on getting better at keeping folks updated here.
Just for easy reference, here’s the December update forum thread and a blog post we published the same day.
Consistently displaying blog posts on the website for all users, not just that subset of users who happen to use iNat’s interface in English might be a good step for doing better at keeping folks updated. It seems like users might be given a choice to decide whether they want to see this or not (say, a profile setting), rather than having someone else assume in advance that they are not interested in certain announcements.
I agree - even an emailed blog post or something would be helpful, and people could turn the option to get that info on and off like notifications. I wouldn’t have seen the voting thing on subspecies and community agreement had I not scrolled all the way down to the bottom of my notifications dashboard. It’s small and easily missed.
On the website it is top right and in your face. It should be visible to all users, I agree.
It depends on how you’re viewing the website. On a phone browser the subspecies trial and other blog posts are at the very bottom of the page.
Proof that “power users” still often miss these is how many curators were unaware of the short-term policy change last year requiring a one month flagging period before committing taxonomic changes for plants. Messaging about policy changes etc is definitely not optimized (even if it feels “in your face” already for you and me).
This is how it shows up on a tablet browser as well.
The notification system in general is really bad. Messages are fine, but if I get tagged in a post I often don’t see them because my inbox is just so spammed with various nonsense notifications that are unfilterable and vanish if I click the wrong button. I just absolutely assaulted with observation field updates because of the sequencing work I do and it makes it extremely difficult to see anything that needs my direct attention.
For what it’s worth, this exists https://www.inaturalist.org/comments?q=@lothlin , it’s not easy to navigate but it does gather them all in one place and is better than nothing
Probably varies a lot from taxa to taxa and by location. At least in North American Lycosidae the user knowledge has improved to the point that more IDs are made. There is also a small “team” of world leaders who have put an incredible amount of effort into cleaning up identifications in the last year or two. Both of these have improved the CV models to the point where they are often correct instead of often wrong. For some new users, a sense of community will come naturally as they start recognizing user names and have special identification problems they have questions about. For others it may not but we still get to see a slice of their natural world.
As far as I’m aware, it does still leave out all comments made within an ID (so will only collect comments posted separately)